Docus
Docus is an AI health platform that turns raw laboratory results into two reports at once: a clinical one for doctors and a plain-language one for patients. Diagnostic labs, physicians and individuals each get a dedicated product.
What is Docus?
Docus is an AI layer that sits between a diagnostic laboratory and the people who read its results. Published by Docus, Inc., a Delaware company founded in 2021, it takes raw test data and returns two documents from a single input: a clinical report for the ordering physician, carrying differential diagnoses, follow-up suggestions and care plans, and a plain-language report that explains the same findings to the patient.
The laboratory product covers routine panels such as blood, urine and biochemistry, infectious disease testing including UTI and STI workflows with resistance markers, pharmacogenomics with drug-gene and optional drug-drug interaction analysis, and anatomic pathology. Interpretation is context-aware: the system draws on the patient's earlier results instead of reading one panel in isolation.
Three modules surround it. Follow-up recommendations suggest the next test and notify patients directly, while doctors can set consultation reminders by hand. A compliance agent checks ordered tests against diagnosis codes to flag medical-necessity problems before a payer denial, verifies repeat-testing frequency, and alerts both the lab and the ordering doctor. Digital requisitions add guided form completion, AI-drafted medical-necessity text, e-signatures and automatic routing into the lab's system.
Two further products share the brand. The Docus Doctor Platform gives independent physicians an AI medical assistant, a research assistant covering literature and clinical guidelines, automatic SOAP notes and patient messaging with AI autocomplete. The AI Health Assistant gives individuals an AI Doctor, lab-result interpretation and long-term memory. A separate human service, Second Opinion, connects users with a claimed 350+ doctors across 30+ specialties for 490 USD per consultation.
Docus documents its own architecture in unusual detail: medical OCR, extraction and retrieval agents built on RAG, a selective memory the user can edit or erase, reference sources including NIH, FDA, PubMed, WHO and ClinicalTrials.gov, a reasoning agent graph, and a quality-assurance agent that scores every answer and regenerates it when the score falls short.
Labs adopt it either through API/HL7 delivery into an existing LIS, CRM or EHR, or through ready-made Docus interfaces that require no integration and can be rebranded with the lab's logo, domain and colours.
What it does
- Turn raw lab results into a clinical interpretation report with differential diagnoses and care plans
- Produce a separate patient-facing report that explains the same results in plain language
- Recommend follow-up tests and notify patients when a repeat or additional test is due
- Check ordered tests against diagnosis codes to flag medical-necessity issues before a denial
- Verify repeat-testing frequency against applicable rules and alert the lab and the ordering doctor
- Collect AI-validated digital requisitions with e-signature and route them straight into the LIS
- Answer clinical or patient questions through a context-aware assistant that remembers prior results
When to use Docus / When not to
A quick filter to help you decide if Docus is the right fit.
When to use Docus
- Reference, direct-to-consumer and hospital-based laboratories that want to ship interpretation alongside their results without building it
- Labs already running a LIS, which can receive Docus output through API/HL7 or through the LigoLab and Lifepoint Informatics partnerships
- Independent physicians handling a large volume of biology, who need differential diagnoses, SOAP notes and follow-up suggestions in one place
- Lab billing and compliance teams exposed to payer denials, thanks to test-versus-ICD checks and repeat-frequency verification
- Clinicians in lab-heavy specialties such as infectious disease, pharmacogenomics, haematology, endocrinology and pathology
When not to use Docus
- Anyone looking for a self-driving diagnosis: Docus states plainly that its AI does not replace a doctor and that the clinician stays responsible
- Buyers who need data residency outside the United States, since processing and storage are declared in the US only
- Procurement teams that require a signed DPA or a published subprocessor list before onboarding, as neither is available on the site
- Users under 21, an age floor set by both the terms of use and the privacy policy
- Radiology and medical imaging workflows, which fall outside the announced coverage of biology, pharmacogenomics and pathology
How to use Docus
A typical end-to-end flow, from setup to results.
- Choose your entry point: labs and clinics start from the demo form, physicians and individuals sign up directly
- For a lab, request a personalised demo through the contact form and describe your volumes and workflows
- Take the free pilot Docus offers laboratories to test the output before committing
- Decide between API/HL7 integration into your existing LIS, CRM or EHR and the ready-made Docus interfaces
- If you use LigoLab or Lifepoint Informatics, rely on the existing partnership rather than a bespoke integration
- Rebrand the Docus interfaces with your logo, domain and colours if you deliver reports to patients
- For a physician or an individual, create an account with no credit card and stay on the free plan while you evaluate
- Submit input as free text describing symptoms and history, or upload a lab report, prescription or scanned document
- Let the OCR and extraction agents read the document, then review what has been stored in the AI memory
- Edit or erase entries in that memory at any time, and change or cancel your plan from the dashboard
Pros & Cons
Pros
- Producing a clinician report and a patient report from the same result set is the core promise, and the site shows previews of both by test family
- The architecture is published in unusual detail, step by step, including the named medical reference sources
- Two very different adoption paths coexist: full API/HL7 integration, or interfaces that need no technical work at all
- Named, dated LIS and EMR partnerships with LigoLab and Lifepoint Informatics, plus a named customer case study
- Permanent free plans on both the doctor and the patient products, with no credit card required
- Physician and patient pricing is published to the cent, and plans can be changed or cancelled from the dashboard
- Personally identifiable information is not required for the core workflows, and the site actively tells users to redact identifiers
Cons
- Laboratory software pricing is not published at all and only reachable through a sales conversation
- No public API documentation exists, even though API/HL7 delivery is a selling point on two product pages
- HIPAA, GDPR and SOC 2 compliance is asserted on five pages but backed by no report, certificate or trust page
- No data processing agreement is published or offered, and no subprocessor list is available
- The vendor never states plainly whether customer data trains its models, and no training opt-out is documented
- Processing and hosting are US-only, with no EU residency option and no Article 27 representative
- The legal pages were last revised in March 2024, before the laboratory offering now front and centre
Pricing & Plans
A permanent free plan is available on both the Docus Doctor Platform and the patient-facing AI Health Assistant, with no credit card required. The lowest paid entry point is USD 3.99 per month for the Patient Lite plan, billed annually at USD 47.88. Physician plans start at USD 25 per month, also billed annually. Laboratory software is priced by quotation only, based on monthly test volume, workflow complexity and regional regulations, and API pricing likewise depends on usage volume and integration requirements. The Second Opinion service is billed separately at USD 490 per consultation.
- Patient Platform Free — USD 0
- 3 messages per week with the AI Doctor
- 1 test result interpreted
- long-term memory
- chat history
- Patient Platform Lite — USD 3.99 per month billed annually (USD 47.88)
- 50 messages per month
- 5 results interpreted per month
- Patient Platform Pro — USD 7.99 per month billed annually (USD 95.88)
- 500 messages per month
- 15 results interpreted per month
- Doctor Platform Free — USD 0
- 1 patient managed
- unlimited AI medical assistant and lab test interpretation
- Doctor Platform Starter — USD 25 per month billed annually (USD 300)
- up to 50 patients
- Doctor Platform Pro — USD 50 per month billed annually (USD 600)
- up to 500 patients
- Doctor Platform Ultimate — USD 100 per month billed annually (USD 1200)
- up to 2000 patients
- Docus Lab Software — flexible subscription quoted on test volume
- workflow complexity and regional regulations
- Second Opinion — USD 490 per consultation
- billed separately from any subscription
Data, GDPR & hosting
A consolidated view of how Docus handles your data.
GDPR overview
GDPR is addressed explicitly. The homepage and three product pages claim security protocols aligned with HIPAA, GDPR and SOC 2, and the privacy policy carries a dedicated "Your GDPR rights" section listing access, rectification, erasure, portability, objection and restriction. Docus describes itself there as a data processor and gives a single contact point, support@docus.ai, for exercising those rights. The wording is hedged: the rights apply "solely to the extent applicable to us". Several expected elements are missing. No Article 27 EU representative is named, no data protection officer is designated, no data processing agreement is published or offered, and no subprocessor list exists. Data is maintained and processed in the United States, with the transfer covered by user consent. No SOC 2 report or certificate backs the compliance claim.
Who owns the data?
Docus, Inc. publishes a privacy policy last revised on 26 March 2024 and describes itself there as a data processor. Users supply personal and medical information voluntarily and may refuse, at the cost of losing certain features. The company states it does not sell, trade or rent user information. Data may be shared with service providers, with medical experts the user selects, with corporate affiliates, and with authorities on legal request. Deletion can be requested by email at support@docus.ai, subject to legal retention duties. Both GDPR rights (access, rectification, erasure, portability, objection and restriction) and HIPAA rights, including amendment, an accounting of disclosures and breach notification, are set out separately.
Reuse rights
Users keep access to their own records and can review, export or delete them, but the terms do not grant any right to reuse Docus output beyond personal or clinical use, and reports remain framed as informational rather than as a formal diagnosis. On the vendor side, the stated purposes are service delivery and improvement, customer support, transaction processing, research subject to separate written authorisation, abuse prevention, and marketing communications in line with the user's preferences. Anonymised findings may be published in journals or on websites, with the commitment that no name or identifying detail is included. Payment card data is handled by PCI-compliant third parties, and Docus says it never accesses card numbers itself.
Data retention & training
Hosting summary
Docus states that it maintains and processes user information in the United States, under US law, and warns that this may not offer the same level of protection as the law of the user's own jurisdiction. Using the service is treated as consent to that transfer. No alternative region is offered, no EU residency option exists, and no hosting provider is named. The front end is served behind Cloudflare, whose anycast address tells you nothing about where data is actually stored. Encryption is claimed "at every step" on the homepage, without technical detail, and the site adds that personally identifiable information is not required for the core workflows. There is no published subprocessor list, so the identity of the vendors involved in storage and processing cannot be checked from the site.
Things to keep in mind
Risks and trade-offs to weigh before adopting Docus.
- Compliance with HIPAA, GDPR and SOC 2 is claimed on five pages but supported by no report, certificate or trust page, so it remains a marketing statement rather than an audited fact
- The vendor never says clearly whether customer data trains its models, and the only related sentence mentions continuously improving OCR models from the data received, with no opt-out documented
- No data processing agreement is published or offered and no subprocessor list exists, which is a real obstacle for any organisation handling health data under a controller obligation
- All processing and hosting happen in the United States, with the transfer covered by user consent and no EU representative designated under Article 27
- Legal pages were last revised in March 2024, before the laboratory offering now promoted on the homepage, so the published terms may not describe what is actually sold today
- Performance figures differ between pages, with +230% and 5% to 27% on the homepage against +200% and a fivefold improvement on the demo form, and neither set is independently verifiable
- Convenience can erode judgement: an AI-written interpretation is easy to accept as read, and both the vendor and common sense insist the clinician stays responsible for the decision
Setup & Integrations
Technical difficulty
It depends entirely on the path. Physicians and individuals sign up on their own, with no technical prerequisite. Laboratories choose between the ready-made Docus interfaces, which need no integration and are advertised as delivering results in minutes, and an API/HL7 connection into an existing LIS, CRM or EHR, which requires engineering time on the lab side. Labs already using LigoLab or Lifepoint Informatics can rely on those partnerships instead. One caveat: no public API documentation exists, so the integration path has to be scoped through a sales conversation.
Deployment
Integrations
Supported languages
Behind Docus
Social
Resources
All the official URLs gathered for verification and reference.
Frequently asked questions
Who is Docus actually built for?
Is there a free plan?
How much does the laboratory software cost?
What types of tests can Docus interpret?
How does Docus connect to an existing lab system?
Where is the data hosted?
Is Docus GDPR compliant?
Is my data used to train the models?
What is the minimum age?
How do I delete my data?
Should you pick Docus?
Docus is a substantial product rather than a landing page. There is an identifiable company behind it, dated legal pages, a postal address and a phone number, published prices on two of the three offerings, report previews by test family, a step-by-step account of how the AI actually works, a named customer case study, and two dated partnerships with LIS and EMR vendors. For a diagnostic lab that wants to deliver interpretation without building it, the proposition is coherent and the two adoption paths, integrated or turnkey, are genuinely different in cost and effort.
The reservations are documentary rather than functional. Compliance with HIPAA, GDPR and SOC 2 is asserted on five separate pages and evidenced on none: there is no report, no certificate and no trust page. No data processing agreement is published or offered, and no subprocessor list exists, although the privacy policy describes six categories of vendors. The company never states plainly whether customer data feeds model training, and documents no opt-out. Hosting is US-only, with no EU representative under Article 27. The legal pages were last revised in March 2024 and predate the laboratory offering the site now leads with, which is a gap worth raising.
None of this makes the tool unserious, but it does set the homework. This is health data supporting clinical decisions, so the documentation bar is higher than for ordinary SaaS. Ask for the SOC 2 report, a DPA and a clear written position on model training before signing anything, and ask why the privacy policy has not followed the product. Physicians and individuals can simply use the free plan and judge the output themselves, which is the cheapest way to find out whether the interpretation quality lives up to the claim.
- Choosing a selection results in a full page refresh.
- Opens in a new window.